Open access peer-reviewed chapter

Advance Intelligent Video Surveillance System (AIVSS): A Future Aspect

By Mritunjay Rai, Agha Asim Husain, Tanmoy Maity and Ravindra Kumar Yadav

Submitted: October 24th 2017Reviewed: March 13th 2018Published: November 5th 2018

DOI: 10.5772/intechopen.76444

Downloaded: 1278

Abstract

Over the last few decades, remarkable infrastructure growths have been noticed in security-related issues throughout the world. So, with increased demand for Security, Video-based Surveillance has become an important area for the research. An Intelligent Video Surveillance system basically censored the performance, happenings, or changing information usually in terms of human beings, vehicles or any other objects from a distance by means of some electronic equipment (usually digital camera). The scopes like prevention, detection, and intervention which have led to the development of real and consistent video surveillance systems are capable of intelligent video processing competencies. In broad terms, advanced video-based surveillance could be described as an intelligent video processing technique designed to assist security personnel’s by providing reliable real-time alerts and to support efficient video analysis for forensic investigations. This chapter deals with the various requirements for designing a robust and reliable video surveillance system. Also, it is discussed the different types of cameras required in different environmental conditions such as indoor and outdoor surveillance. Different modeling schemes are required for designing of efficient surveillance system under various illumination conditions.

Keywords

  • surveillance system
  • AIVSS
  • digital camera
  • types of camera
  • background model
  • illumination

1. Introduction

In recent times, surveillance systems are gaining a lot of popularity. The government, various organizations, residential societies, etc., are using these systems to keep a check on various activities for safety and security purposes. Earlier surveillance systems had a lot of dependence on human operators, it is lately that automated systems are being preferred because of their better efficiencies and reliability [1]. It has been seen that surveillance with full human operators’ involvement has certain inadequacies like the high cost of labor, variations in long-duration capturing and limited ability for multi-screen monitoring [2]. Traditional surveillance systems are being complemented and even replaced by the advanced intelligent surveillance systems (AISS), as the latter is used in identifying abnormal behavior and patterns in videos by developing artificial intelligence technologies, pattern recognition, and computer vision. This enables high accuracy monitoring of more scenarios by a few observers. In the last few years, the video surveillance market has seen a major transformation into third generation video surveillance systems, moving to IP video from traditional analog video causing better processing power and improved compression algorithm [3]. These Intelligent video surveillance systems are not just confined to laboratories but have hit the marketplace as well. With this generation, the era of Intelligent Video Surveillance began, not only in research labs but also in the marketplace. With the start of 2010, many research labs, such as Kiwi Security Labs, started to broadcast the “Advanced Intelligent Video Surveillance Systems” (AIVSS). With this production, a new category of features is presented, which are expected to have a big impact on the marketplace security and a sensor control. The Figure 1, shows an Intelligent Video Surveillance System. All the components of the system are interconnected using many cameras for critical sites, by means of IP mega pixel cameras. Selective ID protection feature has been provided in this architecture of AIVS.

Figure 1.

The distributed architecture of the advanced intelligent video surveillance system.

Here, the disseminated keen design of the AIVS was used to execute the component Selective ID Protection. Appropriately, the system could respect the current security law necessities in a few nations, notwithstanding the prerequisites of governments and knowledge specialists to ensure the character of their operators.

Apart from the hardware (H/W) and software (S/W) which are considered as performance improvisers and the architecture of inter-operational processing, the performance of the system depends upon the privacy of the system. Moreover, in many countries, privacy issues are becoming more crucial and are considered as a performance decelerator. On one side, the performance of the network depends upon the performance of each of Network Element (NE), access performance, transmission performance, etc., which are also considered as performance decelerator and on the other side, the network’s performance depends much upon the Security Management Process [4] of the advanced IVS system. Protection assumes again a noteworthy part of security execution and in security administration forms and accordingly on the system execution as an execution decelerator. From the perspective of the Security Management Process, the suggestive development of process science was the driving potential to build up an elite propelled IVS, which uses a keen Security Management Process, which is controlling the system execution, i.e., system accessibility, secrecy and trustworthiness, bringing about a substantial scale vital answer for security specialists and governments [3, 4], Figure 2 demonstrates the execution effect of the progress smart video surveillance system [5].

Figure 2.

Performance impact on the distributed architecture of the advanced intelligent video surveillance system.

The remaining structure of the article is organized as the Section 2 deals with the basic requirements for designing of video surveillance system including different types of cameras and video management systems using surveillance display. Section 3 discusses the surveillance system for both indoor and outdoor environmental especially with illumination conditions. Section 4 discusses the different modeling schemes used for surveillance systems. Lastly, Section 5 holds the conclusion and the future aspects.

2. Video surveillance system design requirements

This section provides the details of decisions taken while designing the video surveillance system. The design of a video surveillance system requires decisions that need familiarity with the basic options and the basis behind the selection of any available choice in the market [11]. So, designing a system requires better remote access, further remarkable mix with different systems, enhanced picture quality and additionally that requires flexibility with others system [12, 13]. In any case, for end clients to take the full preferred standpoint of the advantages, the outline and execution of the arrangement should be precisely arranged and executed. This will guarantee the system is adaptable and future-sealed and is proper for a client’s need. These six stages cover guidance about choosing the correct hardware, an assessment of the accessible innovation and help with the decisions that should be made. The following decisions are to be made for designing of video surveillance system are as follows:

  1. Camera and its type.

  2. Video management system.

  3. Types of video management system.

  4. Storage type.

  5. Types of video analytics.

  6. Surveillance video display.

2.1. Camera and its type

In late 1990s, the digital cameras came into the market, they were built on Complementary Metal Oxide Semiconductor (CMOS) based image sensor whose performance is better and are cheaper than Charge-Coupled Devices (CCD). It has been seen during the last decade that on an average there is an annual growth rate of around 12% of digital cameras throughout the world. The credit can be attributed built-in intelligent image processing and pattern recognition algorithms. These smart digital cameras can spot motion, detect objects, read vehicle number plates, and even identify human behaviors. They have become an essential component to build active and automated control systems for many applications and will continue to play a significant role in our daily life in the future [7]. Smart cameras are generally intended to perform specific, repetitive, high-speed and high-accuracy jobs. The typical applications of these smart cameras are Machine vision or intelligent video surveillance systems (IVSS).

Video surveillance technology is functionally used in traffic cameras [9], which are used for traffic footage recording and are many times shown during traffic reports on TV news. They are placed over the traffic signals, along with the busy roads, and at busy junctures of the highway. Whether they record the movement of traffic for future study or to monitor traffic and issue challans/tickets for any traffic rule violations, they are an extremely popular form of video surveillance. They are commonly used in the monitoring & management of traffic, computerized parking garages, driver support and control access systems, etc. License Plate Recognition (LPR) is the most well-known and widely used application in the category of traffic management and monitoring. Yet, due to increasing demand other categories of vehicle classification have been added recently. Make and Model Recognition (MMR) & Color Recognition (CR) of cars is a major and comparatively new functionality which helps in detecting the model of the car, along with the vehicle types for, e.g., Light Motor Vehicle, Heavy Motor Vehicle, etc. Installation of Camera plays an important role in the advance intelligent video surveillance system.

Cameras are the key contributors to the video surveillance system. The camera position and the type of cameras used under various conditions are important factors in video surveillance. These two parameters are briefly explained below:

  1. Positions for camera installation: Cameras should be placed in appropriate areas to record relevant video. The appropriate areas for proper placement of cameras can be entrances, hallways, driveways, T- Points, highway intersection points, exits, etc., and in areas where there is a high density of people or vehicles. Moreover, cameras can be placed in areas that require security such as parking spots, VIP areas, schools, restaurants & hotels, bank locker rooms, hospitals, etc. Planting cameras at crucial and suitable points is a cost-effective way to monitor and document people and vehicles arriving and departing certain facility.

  2. Type of cameras to be used: There are many types of camera available on the market. The suitability of the camera depends upon the situation in hand. Fixed camera can be used for recording only one specific view while a PTZ camera is generally used to cover wider fields of views. Mostly fixed cameras are used in video surveillance as they are five to eight times less costly than PTZ cameras. Color cameras are preferred during day time and in highly illuminated areas. However, during night time and in poorly lit areas infrared or thermal cameras are used that gives black and white images. Thermal cameras can also be used under settings of complete darkness, where they produce only contours of objects. Cameras can be standard definition or high definition cameras that provide a resolution of up to 16 MP.

IP cameras digitize the recordings within the camera while analog cameras’ recordings (which are used as surveillance cameras) are digitized on the computer. Video surveillance systems usually make use of a combination of different type of cameras. Some of the camera types are discussed briefly as under:

  1. PTZ camera: One of the commonly used camera for security purpose is PTZ camera; where P stands for Pan, T for Tilt and Z for Zoom. Pan, Tilt, and Zoom are the main features of this camera which is controlled by a software or via joystick. This security camera has an ability to rotate 360 degrees so that it can cover a wide area and can zoom into detail. The other features that attract toward this security camera are Weatherproof, Night Vision, Multiple Alarms, Auto Focus, and Tamper Resistant.

  2. Box camera: Box Style security camera is an outdoor camera where customization of the lens is possible. The lens can be variable or fixed. Box surveillance camera is an ultra-high-resolution CCTV camera made with the new image sensor processor which is capable of capturing video at 700 TV lines of resolution in color and black & white, 960H CCTV resolution. This box camera includes a 6-60 mm variable focal auto-iris lens which gives security installers a lot of flexibility to adjust the camera angle of view and zoom level.

  3. Dome camera: It is a combination of lens, camera and ceiling mount packaged in a dome shape. This is well suited for surroundings that tend to get dirty, like kitchens and storerooms, etc., the best part of it is compact in size and artistically very attractive too.

  4. IP camera: An Internet Protocol camera generally transmits a digital signal using Internet Protocol over a network. The main features of these security cameras are its high resolution and scalability. Right now, up to 30 Mega pixels are available in the market.

  5. Wireless IP camera: As its name suggests that this type of camera is completely wireless, installation is easy and reduces the rate of network cabling. The camera also has the feature of tilting and revolving which helps in maximum viewing with clarity and even in low light conditions.

  6. Bullet camera: This security camera shaped like a bullet which is a combination of camera, lens and packaged in a bullet style. This camera is good for dim light situations and can be easily mounted to ceilings or walls because most of them use a tri-axis type of base. Bullet cameras come in all sizes (small, medium & large). Infrared bullet cameras generally are larger in diameter to put up the extra space that their infrared Light Emitting Diodes require.

  7. Day and night camera: This security camera is used for both indoor and outdoor environments with low or dim illumination conditions. A day and night camera has distinctive lenses that permit infrared emission formed by infrared LEDs and imitated from objects to go through and reach to a Charge Coupled Device or Complementary MOSFET chip inside the camera. As a result, the end user can see the picture in total darkness at the distance of infrared emission produced by LEDs. A day and night camera can have infrared LEDs mounted on its housing or can accept the emission, produced by an infrared steeple. A Day and night cameras over and over again have changes in their digital signal processor that pays for the alteration in illumination between day and night methods.

  8. Thermal camera (FLIR): The first commercial thermal imaging camera was used in 1965 for high voltage power line inspections. Since then the utility of thermal imaging cameras for industrial applications has become a pivotal market segment for FLIR (Forward-looking IR) systems, a later name for high voltage power lines. The thermal imaging technology has drastically evolved since then, and thermal imaging cameras have evolved to become compact in size and look like a digital photo camera, they are now easy to use and produce real-time crisp high-resolution images making them a widely important tool for industrial applications [8]. They can detect anomalies that are generally invisible to the naked human eye, thus taking corrective preventing costly systems going for a total breakdown. Thermal imaging cameras are used to determine the maintenance requirements for electrical and mechanical installations as they tend to generate unusual heat before they fail. Preventive actions can be taken by discovering these hot-spots. A thermal imaging camera is a non-invasive instrument which scans and visualizes the temperature distribution of surfaces of a machine quickly and accurately, thus reducing cost and saving time across the world.

2.2. Video management system

Video management system is the recording and management of access to the video, which is captured by a camera and is then transferred to the module of the video surveillance system [4]. There are two types of connections through which the captured video is transferred:

  1. Videos can be transmitted over the computer network IP or they can be sent as analog videos. Videos from both IP cameras and analog cameras can be transferred over the computer network whereas unlike analog cameras, IP cameras can connect directly to an IP network. In case of analog cameras, an encoder must be installed to transmit analog video over IP. The input from an analog camera is encoded and output a digital stream for transmission over an IP network.

  2. Depending upon whether IP camera or analog video camera is used, the captured video can be transmitted over cables or through the air. Cables are generally considered inexpensive and the most reliable method of transferring video but, wireless is an important alternative for transmitting videos as setting wires can be expensive for certain applications such as parking lots, fence lines, remote buildings, etc.

2.3. Types of video management system

In a Video management system, videos taken by the cameras are stored, managed and are transmitted to various viewers. The video management systems usually used in video surveillance systems are:

  1. In a digital video recorder (DVR), videos are recorded from a surveillance camera on a hard disk. It is a security system device in which the rate of the frame can be converted from real-time to time lapse to save the disk space. They are more flexible as compared to earlier analog VHS tape systems and allow easier transmission of video over a computer network. Digital Video Recorders accepts only analog camera feeds as inputs and supports remote viewing over the Internet. DVR is a combination of software, hardware, and video storage.

  2. Hybrid digital video recorders (HDVRs) support IP cameras. They can perform all the functions of a digital video recorder mentioned above and adds support for IP and megapixel cameras.

  3. Network video recorder (NVR) supports IP cameras only, however, to support analog cameras it requires an encoder. NVR can record videos from a no. of digital CCTV cameras that are transmitted over the network.

  4. IP video surveillance software is a product application that does not accompany any equipment or capacity. The client must load and set up the PC/server for the product which gives considerably more prominent opportunity and possibly bring down cost yet in the meantime it accompanies noteworthy greater many-sided quality and time important to set up and advance the system. IP video surveillance software is the most regular decision for video systems that contain extensive camera tallies like at least hundreds.

2.4. Storage type

In a video surveillance system, storage of the surveillance video is very vital. This video is used for later retrieval and review. Cost of storage and security related fears specific to the application of the video surveillance system determines the duration for which the video should be stored [11]. For example, in supermarkets and restaurants video recordings are kept for a relatively shorter duration as compared to the bank where there is a greater need to hold videos for a longer duration (60–90 days) as there is a major threat of fraudulent investigations that are often reported after many days of the incident. The digital data is stored permanently in the Storage, till it is purposely deleted. Even without power, this source holds its content. Storage generally means magnetic disks, solid-state disks, and USB drives and may also refer to magnetic tapes and optical discs like CDs, DVDs, etc. Although storage prices are falling, the demand for the surveillance system and for the amount of storage is rising. Several techniques have been developed to optimize the use of storage because of its high cost. There are three main types of storage:

  1. Hard drives that are built inside a digital video recorder, network video recorder or server represents the internal storage. It is the most reasonably priced but may be less reliable and scalable. Most frequently it is used in video surveillance systems and can provide a storage of 2 TB to 4 TB.

  2. Directly attached storage are the hard drives that are located outside of the digital video recorder, network video recorder or server. It is more expensive as compared to internal storage but has greater scalability, flexibility, and redundancy.

  3. Capacity clusters are IP based capacity places had some expertise in putting away video gushing from an expansive number of cameras. They give proficient, adaptable and versatile capacity.

2.5. Video analytics type

Video analytics encompasses the below-mentioned tasks:

  1. Storage optimization: Storage optimization is realized based on the motion detection. The video management systems decide to store the video when any motion/ moving object [10] is spotted in the observed scene or else the video is either not stored or is stored at a lower frame rate or a lower resolution to save storage space. Cameras may capture long durations of inactivity when placed in buildings when they are locked, staircases, etc. This application helps in reducing the consumption of storage by 60–80% as compared to continuous recording.

  2. Identify threatening events: Video analytics can also be used to identify threatening events to pro-actively identify any lapse in security incidents, be alert, and to stop them; for example, license plate recognition, perimeter violation, abandoned objects detection, and people counting.

2.6. Display of surveillance video

Videos captured by a surveillance system are eventually viewed by human beings and is usually used for past investigations. Some surveillance videos are watched online continuously, e.g., in educational institutions to keep a check on student actions, in shops to keep an eye on shoplifters and in public areas to identify criminal threats. Some surveillance videos are viewed online infrequently by the owner of the apartment. Videos can be viewed in 4 different ways:

  1. Local: It is viewed directly from the digital video recorder. Small facilities like Banks, retail outlets, and small businesses ideally use the network video recorder to monitor their sites.

  2. Remote: It is viewed through standard remote PCs for viewing live and recorded videos through an installed application, a web browser, or a powerful web viewing.

  3. Mobile: This kind of viewing allows an instant check of the captured video. It holds great importance in video surveillance systems. Mobile clients exist in the market for the last few years, but there are challenges related to its implementation on PDAs/phones. However, a few latest technology phones have renewed interest in mobile viewing.

  4. Video: Big security operation centers where a lot of cameras must be examined or scrutinized, video wall viewing is generally preferred. Video walls offer a very big screen so that many people can watch the captured videos from a number of cameras at the same time. They can change between numerous video streams and could automatically show videos from points where alarms have been triggered.

3. Surveillance system and its types

The word Surveillance has been derived from the French word “sur” means “from above” and “veiller” means “to watch.” Surveillance means to monitor behaviors, movements, activities, and information for controlling, managing, and protecting people. It can include observing from assistance through an electronic device like CCTV cameras (Closed-circuit television) or by keeping a track on electronically transmitted information like on phone calls & internet traffic. It may also include a number of or relatively lesser technology means such as intelligence agents, detectives, etc. Surveillance systems are readily being used by governments for crime prevention and investigation, in intelligence gathering, and to protect people, objects, processes, etc. For many, surveillance may be a violation of one’s privacy and has often been criticized by many civil liberty activists. Laws in many countries have restricted their domestic government & the private use of surveillance, generally restraining it to situations where public safety is in jeopardy.

Dictator government at times have any residential confinements, and universal secret activities are normal among a wide range of nations. While observation systems have been exceptionally ordinary for business properties, it took a while for them to wind up plainly mainstream for private homes too. One reason is that wrongdoing insights demonstrate that the further developed and cutting edge a security and observation system is, the more culprits will maintain a strategic distance from them through and through. Also, the cost of camera gear for home utilization has altogether dropped as of late. The best sorts of observation system have certain traits that you should give careful consideration to remember the ultimate objective to ensure that you totally secure your property. One fundamental thing is that the system must be effortlessly expandable to guarantee that as and when required you can cover more indoor and outdoor regions with cameras.

More established systems can be extremely restricted once introduced and will likewise just permit a specific constrained measure of identification gadgets, including cameras, vibration, and movement locators. This can turn out to be expensive if a system must be supplanted because of extending business or private premises. Here are two or three signs on what to pay uncommon identity to while exploring diverse sorts of security systems that will expand insurance.

3.1. Outdoor video surveillance

Outdoor video cameras play a very important role in law enforcement by not only capturing video of potential criminals but also by preventing crimes. Past statistics on crimes show that the more noticeable and better technology the camera systems are, the more it helps in preventing criminal activities in business or residential buildings [6]. Factors that are necessary for an outdoor camera are: they should be weather- resistant, and should include night vision even in well-lit locations. This helps in ensuring that switching off the light may not affect the camera captures. Cameras should be installed at a point capturing a wider angle & that is not easily accessible from the ground.

3.2. Indoor video surveillance

There is no necessity for Indoor cameras to be weatherproof, that means they are easier to install, are smaller with lesser restrictions. For indoor locations as shown in Table 1, a camera that provides a very wide angle of view is needed so as reduce the requirement of the no of cameras and to reduce the dark spots. Night vision and quality of the video captured are also essential for Indoor video surveillance.

Table 1.

Different application areas of video surveillance system.

3.3. Illumination and artificial lighting

Amid no light conditions, it is not conceivable to see anything, yet to security cameras, they are barely exceptional, some high fragile security cameras can get clear monochrome images in starlight lighting up condition, they in like manner can see the object in whole darkness while using additional infrared lighting. In this section, we offer the essential finding out about the general lighting up, and the wide grouping of fake lighting. Lux is a prevented unit in see from claiming lumen, and the lumen is a precluded unit in light from securing candela. The lumen (structure: lm) is the SI unit of luminous change, a measure of the vitality of light obvious by the human eye and the candela is the SI base unit of luminous intensity. One lux is equal to one lumen for each square meter, where 4n lumens is the total luminous change of a light wellspring of one candela of luminous intensity.While picking a proper surveillance camera to present, make a point to consider the illumination condition in the environment. Lux, it is the illumination level unit used to address the allocated domain illumination. In incredible illumination region, the customer can use a general execution security camera. In any case, the shading system for environmental illumination is under 2.0 Lux, a monochrome illumination system for under 0.2 Lux environment use the higher execution security camera (i.e., starlight security camera), which is extremely fundamental. Using Lux meter can evaluate illumination level. If we do not have a lux meter, we can follow the general illumination data table.

Utilizing infrared LEDs to transmit vague (to human) infrared lights. The infrared light wave length is 850 nm, which empowers camera to get monochrome images. While using the IR illumination, the camera will encounter infrared-submersion issue because the photo setback purposes of enthusiasm for objects arranged in central and short division watching an area. Remembering the true objective to handle this issue, IR sharp development was brought into various security cameras. The Infrared splendid limit can modify camera’s Infrared intensity as demonstrated by the watching objects, keep up a vital separation from IR-drenching issue. Starting at as of late, the IR illumination can cover 0–200 meters independent. Using white light LEDs to illuminate the area under observation. The white light LED wavelength is 450 nm, which has a place with noticeable light.

The white light illumination can empower camera catch shading images in low illumination or zero illumination environments. Compared with infrared illumination, white light illumination can work in particular application, for example, acknowledgment of vehicle number plate. Moreover, the white light illumination can be utilized to stop interlopers/crooks. Sony double light IP camera can naturally turn on white light illumination when individuals stroll into observing territory. The white light has substantially shorter illumination separate than infrared, its range is 0–50 m. Frequently, laser maker is set up into PTZ camera which offers 30× optical zoom limit. The bigger piece of laser maker utilizes 808 nm wavelength diodes, laser maker has various purposes of intrigue; long detachment illumination, adaptability, acclimate to the environment, long life expectancy. Laser illumination can help the camera to get clear images with high clearness. Besides, it enables the camera to get pictures inside 1 kilometer or even 3 km long partitioned.

In indoor surveillance, the cutting edge highlights liberal change revelation and following estimation. Indoor conditions are passed on utilizing unmistakable truly little spaces that are pulled back with dividers and give each other through passages and sections. In this condition, it is crucial to relate the district of a specific individual in various parts of the building structure. It is less pivotal to track continually the difference in a man as this movement will no vulnerabilities or conceivably buts break reliably because of building’s geology. In outside surveillance, the cutting edge moreover merges veritable change request and following tallies. The limit is that these figures are normally stunningly besides made than the relating indoor checks on account of the distinctive quality presented by exceedingly factor lighting.

The topology of outside conditions is correspondingly completely not exactly the same as that of indoor conditions. Moving things are people and what more vehicles is in like manner, going at all around higher speed. Snappier moving things require faster sorting out paces, yet the figuring is liberally more computationally than those related to indoor surveillance conditions. These repudiating necessities on a to a great degree basic level mean the specific difficulties of a pushed outside security system. Utilizing Lux meter can check light level. If you do not have a lux meter, you can propose Tables 2 and 3 to the running with general illumination information.

S. No.PlacesLuminance intensity (in Lux)
1.Warehouse20—75
2.Emergency passway30—75
3.Corridor75—200
4.Shop75—300
5.Office300—500
6.Bank200—1000
7.Meeting room300—1000

Table 2.

Various indoor illumination conditions at different places.

S. No.PlacesLuminance intensity (in Lux)
1.Sunny10,000—1,000,000
2.Cloudy100—10,000
3.Dawn Twilight1—10
4.Full moon over head0.1—1
5.Quarter moon0.01—0.1
6.Sunny Starlight0.001—0.01
7.Cloudy Starlight0.0001—0.001

Table 3.

Various outdoor illumination conditions at different places.

At the point when security camera works in entire darkness environment (i.e., 0 Lux), the camera picture sensor would not have the capacity to catch images. In this condition, the cameras have an artificial lighting system, for example, infrared LEDs, white light LEDs, and also laser producers. These counterfeit lighting systems can help the camera to catch clear monochrome or shading images in low illumination or zero illumination environments. As of late, Hikvision and Sony propelled day and night security cameras which use double lighting system.

4. Different strategies for smart video observation

The present security system could be outlined as takes after: (a) security system act locally and they do not participate in compelling way (b) high esteem resources are ensured deficiently by obsolete innovation system. (c) Reliance on escalated human focus to identify and survey dangers. Various strategies and calculations have been created and actualized, basically in programming, for question following, identification, and acknowledgment. A couple of endeavors have been made to execute a portion of the calculations in equipment. Be that as it may, those endeavors have not yielded ideal outcomes as far as exactness, power and memory necessities. The decision of the ideal calculation can upgrade the execution and help in settling these difficulties. Many question discovery calculations are by all accounts fantastic applicants (e.g., difference-of-Gaussians (DoG), maximally stable extremal regions (MSER), fully affine invariant feature detector (FIAF), scale invariant feature transform (SIFT), speeded up robust features (SURF), background subtraction, and so on.), contrasting in their capacities and prerequisites.

  1. Background subtraction

Foundation subtraction is generally utilized for recognizing moving articles from static cameras. By evaluating the foundation, it would then be able to subtract it from the info outline, by applying some limit esteem, to get the frontal area, i.e., the question. Diverse procedures could be utilized to gauge the foundation, the easiest expect the foundation to be the past casing, and another probability is to apply a mean/middle channel for the keep going N outlines, and accepting the foundation to be the outcome. This calculation is versatile to dynamic foundation changes, simple to actualize and quick and pertinent for constant usage. Be that as it may, the disadvantages are its reliance on the question speed, outline rate, colossal memory, and above all, the edge utilized is neither worldwide nor time-invariant. A vigorous foundation subtraction calculation ought to have the capacity to deal with lighting changes, monotonous movements from the mess and long-haul scene changes [14, 15]. The accompanying examinations influence utilization of the capacity of V(x,y,t) as a video to succession where t is the time measurement, x and y are the pixel area factors. For example, V(1,2,3) is the pixel power at (1,2) pixel area of the picture at t = 3 in the video grouping. A portion of the foundation subtraction techniques are examined underneath:

1. Using frame differencing

A movement identification calculation starts with the division part where the frontal area or moving items are sectioned from the foundation. The easiest method to execute this is to take a picture as foundation and take the casings got at the time t, indicated by I (t) to contrast and the foundation picture meant by B. Here utilizing basic number-crunching computations, we can portion out the articles basically by utilizing picture subtraction method of PC vision importance for every pixel in I(t), take the pixel esteem indicated by P[I(t)] and subtract it with the comparing pixels at a similar position on the foundation picture meant as P[B].

In a mathematical equation, it is written as:

P[F(t)] =P[I(t)]-P[B].

The background is assumed to be the frame at time t. This difference image would only show some intensity for the pixel locations which have changed in the two frames. Though we have seemingly removed the background, this approach will only work for cases where all foreground pixels are moving and all background pixels are static. A threshold “Threshold” is put on this difference image to improve the subtraction (see Image thresholding).

|P[F(t)]E[F(t + 1)]| > Threshold

This implies the distinction picture’s pixels’ intensity is “thresholded” or sifted based on the estimation of Threshold. The precision of this approach is reliant on speed of development in the scene. Quicker developments may require higher edges

2. Mean filter

For figuring the picture containing just the foundation, a progression of going before pictures arrive at the midpoint of. For figuring the foundation picture now t,

B(x, y, t) =1Ni=1NV(x,y,ti)

where N is the quantity of going before pictures taken for averaging. This averaging alludes to averaging comparing pixels in the given pictures. N would rely on the video speed (number of pictures every second in the video) and the measure of development in the video. In the wake of figuring the foundation B(x, y, t) we would then be able to subtract it from the picture V(x, y, t) at time t = t and limit it. In this manner the closer view is given as:

|V(x, y, t)-B(x, y, t)| > {Th}

where Th is a threshold. Similarly, we can also use median instead of mean in the above calculation of B(x, y, t).

a. Maximally stable extremal regions (MSER)

The MSER calculation is an intrigue area identifier initially utilized as a part of wide-standard stereo coordinating. MSER works on the information picture straightforwardly with no smoothing, which brings about the location of both fine and coarse structures [16]. MSER performs all around contrasted with other nearby finders. The principle favorable circumstances of the MSER recognition are that it is the speediest relative invariant area locator. To the best of our insight, the main downside of the MSER is that its execution debases with obscured pictures, which can be settled utilizing keen establishment of the camera topology.

b. Speeded up robust features

SURF is a scale-and pivot invariant intrigue point indicator and descriptor. The calculation extricates striking focuses on the picture and registers descriptors of their surroundings that are invariant to scale, turn and brightening changes. Nonetheless, identification and extraction are computationally requesting and consequently cannot be utilized as a part of systems with restricted computational power [17].

5. Discussion and conclusion

The principal points of Advance Intelligent video surveillance system (AIVSS) are to build up an observation system which can function as an indoor/open-air observation system. As Advance Intelligent Video Surveillance System has a more extensive degree to take a shot at. As nowadays security and protection assume an essential part of the survival of the individual. The perfect observation engineering will have the accompanying attributes: elite, adaptability, simple upgradability, low advancement cost, and a movement way to bring down cost as the application develops and volume inclines. Also, the step by step expanding innovations restricted the working of the Surveillance system, therefore the level of security must be expanded with a specific end goal to stop the obstruction of interlopers. At present, the video surveillance industry utilizes simple CCTV cameras and interfaces as the premise of observation systems. These system parts are not effortlessly expandable and have low video determination with practically zero flag preparing. Nonetheless, the up and coming age of video surveillance systems will supplant these segments with more current computerized LAN cameras, complex picture handling, and video-over-IP steering. They will never again be essentially surveillance camera systems yet in addition video correspondence systems.

The internet protocol (IP) based structure of the new surveillance systems takes into consideration versatility, adaptability, and digital security. Different encoding and translating gauges transport the video stream (MPEG4 CODEC is the standard utilized today). Other than the CODEC work, picture pre-and post-handling improves the photo quality progressively with low dormancy. Programmable rationale with inserted DSP squares, recollections, interfaces, and off-the-rack IP arrangements enables a planner to meet the new system requirements. Security surveillance systems can be generally isolated into a few components, for example, cameras, interchanges, stockpiling, picture preparing, and administration and back-end. Beginning with the camera, the present observation cameras are pushing toward the top-quality period. Regardless of whether it is an IP camera that has turned out to be generally acknowledged or HD-SDI cameras broadcasting in superior quality, they both can give up to full 1080p HD determination surveillance pictures. Giving completely clear pictures is the main role of shrewd surveillance, so top quality picture catch is fundamental, so the data gave by the camera can be handled precisely. In this way, top notch cameras have turned into a key part of observation system merchants. Be that as it may, cameras are only one a player in the general surveillance system, and regardless of whether an ever-increasing number of cameras and video encoders are incorporated, progressed and complex picture examination is still performed on the backend, regularly using cloud-based preparing administrations.

How to cite and reference

Link to this chapter Copy to clipboard

Cite this chapter Copy to clipboard

Mritunjay Rai, Agha Asim Husain, Tanmoy Maity and Ravindra Kumar Yadav (November 5th 2018). Advance Intelligent Video Surveillance System (AIVSS): A Future Aspect, Intelligent Video Surveillance, António J. R. Neves, IntechOpen, DOI: 10.5772/intechopen.76444. Available from:

chapter statistics

1278total chapter downloads

More statistics for editors and authors

Login to your personal dashboard for more detailed statistics on your publications.

Access personal reporting

Related Content

This Book

Next chapter

Real-Time Action Recognition Using Multi-level Action Descriptor and DNN

By Cheng-Bin Jin, Trung Dung Do, Mingjie Liu and Hakil Kim

Related Book

First chapter

A Personal Robot as an Improvement to the Customers’ In- Store Experience

By Joana Santos, Daniel Campos, Fábio Duarte, Filipe Pereira, Inês Domingues, Joana Santos, João Leão, José Xavier, Luís de Matos, Manuel Camarneiro, Marcelo Penas, Maria Miranda, Ricardo Morais, Ricardo Silva and Tiago Esteves

We are IntechOpen, the world's leading publisher of Open Access books. Built by scientists, for scientists. Our readership spans scientists, professors, researchers, librarians, and students, as well as business professionals. We share our knowledge and peer-reveiwed research papers with libraries, scientific and engineering societies, and also work with corporate R&D departments and government entities.

More About Us